Agents are the next big thing after AI. While some people define these “agents” differently than others, the general idea is AI-driven tools that allow tasks to be autonomously executed.
Agent hype reached fever pitch, but one startup was relatively early in the game: Llamaindex. Founded in 2023 by former Uber research scientists Jerry Liu and Simon Suo, LlamainDex allows developers to build custom agents with unstructured data.
“Llamaindex began in November 2022 as an open source toy project,” Liu told TechCrunch. “I was deeply interested in understanding how large language models (LLM) can be used on top of their own data outside of the training set, and I built the first toolset where developers can index and include data in LLM apps.”
With Llamaindex's open source software, which has won millions of downloads on Github, developers can create custom agents that can extract information, generate reports and insights, and perform specific actions. LlamainDex offers data connectors and utilities such as Llamaparse. This converts unstructured data into a structured format that can be used for specific AI applications.
There are other open source frameworks that build AI agents, but LlamainDex is distinguished by its suite of data intake, data management, data indexing and search solutions. Connect data from files such as PDFs and PowerPoint presentations, as well as apps such as concepts and Slack with agents.
Salesforce, KPMG and Carlyle are among the companies that use Llamaindex today, Liu said.
Llamaindex co-founder Jerry Liu.imageCredit headshot: Llamaindex
“Although all of these competing solutions solve specific problems in different parts of the generated AI stack, it is the developer's responsibility to connect fragmented solutions to create work agents,” added Liu. “This is an important issue that hinders transportation agents to production. Llamaindex has made it our mission to provide the safest, accurate and easy to use platform for building end-to-end knowledge agents.”
The next chapter of Llamaindex is an enterprise service built on the company's open source offering. Called llamacloud, customers can create cloud host agents that can manipulate and manipulate unstructured data in a variety of formats.
LlamaCloud can be deployed in a virtual private cloud or installed software as a service, and comes with features such as role-based access control and single sign-on.
To fund the development of LlamaCloud, Llamaindex recently raised $19 million in a Series A funding round led by Norwest Venture Partners, and also saw participation from Greylock. The new cash will raise the total funds for Llamaindex to $27.5 million, and Liu says it will be used to expand its 20-person team and product development.
“There's enough runways to take us through the initial commercial expansion of the platform,” Liu said. “We are betting on a future where developers play a major role in delivering Genai applications within the enterprise.”